LH and FSH, luteinizing and you can hair follicle-revitalizing hormones, respectively; SHBG, gender hormones-binding globulin; IGF-We, insulin-eg growth foundation We
Bad experiences, blood matters and you may chemistries, PSA, plasma lipids, full and you can free testosterone, luteinizing hormone (LH), sex steroid-joining globulin (SHBG), and you may insulin-such as gains grounds I (IGF-I) account was basically measured sometimes during the manage and you may medication attacks. Serum overall testosterone is actually counted of the an immunoassay (8-11); free testosterone of the equilibrium dialysis (43); LH, SHBG, and PSA by immunoradiometric assays (9-11); and you can IGF-We by acid-ethanol removal and immunoassay (28). The sensitivities and you may intra- and interassay coefficients away from adaptation to have hormonal assays were the following: overall testosterone (0.six ng/dl), 8.dos and you will 13.2%; 100 % free testosterone (0.22 pg/ml), cuatro.dos and you will twelve.3%; LH (0.05 You/l), ten.seven and thirteen.0%; SHBG (six.twenty-five nmol/l), 4 and you can six%; PSA (0.01 ng/ml), 5.0 and you can six.4%; and you will IGF-I (80 ng/ml), cuatro and you can six%, respectively. Such assays was validated in earlier times (8-11).
All variables were examined for their distribution characteristics. Variables not meeting the assumption of a normal distribution were log-transformed and retested. An ANOVA was used to compare change from baseline in outcome measures among the five groups. All outcome measures were analyzed using a paired t-test to detect a nonzero change from baseline within each group. P < 0.05 was considered statistically significant.
To explain the relationship ranging from testosterone levels (T) and alter from inside the body weight-free bulk (Y) throughout the testosterone administration, we examined around three patterns: an effective linear design (Y = a great +bT); good logarithmic model, Y = an effective +b · X, where X = record (T), and you may a great andb show the intercept and you can slope, respectively; and you can a growth model, Y = a/(1 +b · age ?k · X ).More